Original data · 5,782 UK CV scans · March–June 2026
The “75% of CVs are rejected by ATS” statistic is a myth. We measured what actually happens.
Almost every careers article repeats the same claim: three quarters of CVs are binned by applicant tracking systems before a human sees them. Nobody cites a source, because there isn't one. We analysed 5,782 real CV scans against real UK job descriptions — here is what the data actually shows.
score below the 75/100 interview-ready threshold
score under 50 — at genuine risk of automatic filtering
average keyword match between a CV and the advert it targets
keywords missing from the average CV vs its target job
How 5,782 CVs actually scored
Every scan compares one CV against one specific job description and produces a 0–100 ATS score. A score of 75+ means the CV is likely to pass automated screening for that role; under 50 means keyword and structure gaps serious enough to risk automatic rejection.
Likely to pass automated screening for the role scanned against
Parseable, but missing keywords and section signals recruiters filter on
Serious keyword and structure gaps — at real risk of automatic rejection
The honest headline: roughly 1 in 3 CVs falls below the interview-ready threshold — a real problem, but half the size of the one the industry keeps quoting. The catastrophic-failure rate is 1 in 8, not 3 in 4.
The real problem: a 48% keyword match
Among CVs that went on to full optimisation (248 deep analyses), the average CV matched only 47.8% of the keywords in the job advert it was written for, missing 9.1 keywords on average. And the missing keywords are not the ones people expect:
| Most-missed keyword | CVs missing it |
|---|---|
| Continuous improvement | 8 |
| Compliance | 8 |
| Customer service | 7 |
| SLA / service levels | 6 |
| Change management | 6 |
| Stakeholder management | 5 |
| KPIs | 5 |
| Training | 5 |
| Agile | 5 |
| Market research | 5 |
Not a single programming language or technical qualification tops the list. The keywords costing people interviews are process and governance terms — continuous improvement, compliance, stakeholder management, SLAs, KPIs. Candidates write about what they did; job adverts screen for how they worked. That mismatch is invisible when you proofread your own CV, which is exactly why it survives application after application.
Methodology & limitations
- Sample: 5,782 scored CV scans on JobSpace AI between 1 March and 11 June 2026, each against a specific job description (2,200+ distinct job titles). Built-in example jobs and internal test scans are excluded.
- Scoring: our ATS analysis model scores keyword match, section structure and parseability on a 0–100 scale. Scores cluster on round values, so we publish banded figures rather than implying single-point precision.
- Keyword depth stats: drawn from the 248 CVs that went through full optimisation — a subsample that skews toward CVs that needed work. Averages from it are labelled accordingly.
- Bias: people who scan their CV suspect it needs checking. If anything, the true below-threshold rate across all UK job seekers is likely lower than 35% — which makes the gap to the quoted “75%” even wider.
Citing these statistics
These figures are free to cite under CC BY 4.0 — please attribute by linking to this page. Suggested citation: “JobSpace AI analysis of 5,782 UK CV scans, March–June 2026”. We re-pull these numbers periodically; this page always carries the latest figures and the date they were updated.
What this means for your CV
- Ignore the 75% scare stat — but take the 35% seriously. You don't need luck to pass screening; you need a CV written against the specific advert.
- Mirror the advert's exact wording. A 48% keyword match means half the advert's language is absent from the average CV. ATS systems match terms, not synonyms — see our guide to passing UK ATS screening.
- Add the process words you actually earned. If you ran stand-ups, handled complaints to SLA, or reported against KPIs, say so in those words — they are the most-screened, most-missed terms in the data.
- Formatting still gates everything. Keywords can't match if the parser can't read them — single column, standard headings, ATS-safe formatting.
Where does your CV sit in this distribution?
Free scan against any UK job description — instant score, no sign-up required.
Check My CV FreeFrequently asked questions
What percentage of CVs are actually rejected by ATS in the UK?
In our dataset of 5,782 CV scans against real UK job descriptions (March–June 2026), 35.4% scored below the 75/100 interview-readiness threshold and 13% scored below 50, the range where automatic filtering is most likely. The widely quoted claim that "75% of CVs are rejected by ATS" has no traceable primary source and is roughly double what we measure.
Where does the "75% of CVs rejected by ATS" statistic come from?
It is usually attributed to a vendor survey from around 2012, but no published methodology exists and the figure predates modern ATS parsing. It has been repeated between careers blogs for over a decade without primary data behind it.
What keywords are CVs most commonly missing?
Across full CV analyses the most-missed keywords were not technical skills but process and governance terms: continuous improvement, compliance, customer service, SLAs, change management, stakeholder management and KPIs. Candidates describe what they did, while job adverts screen for how they worked.
How were these statistics collected?
Each data point is one CV scanned against a specific job description on JobSpace AI between 1 March and 11 June 2026. Scores are produced by our ATS analysis model on a 0–100 scale and reported in bands. Built-in example jobs and internal test scans are excluded. The sample is self-selected (people checking their own CVs), so it may skew toward candidates who already suspect a problem.